AI Recruitment Trends: Adoption & Human Oversight
How AI tooling is being adopted across the recruitment lifecycle — and why human oversight remains central.
Recruitment teams are adopting AI tooling unevenly — heavily in some parts of the process, cautiously in others, which reflects where the risk of a wrong call is highest. Surveys by LinkedIn (Future of Recruiting 2025) and SHRM (2025 Talent Trends) both report adoption rising year on year.
Higher adoption
Resume parsing, job description generation, candidate matching suggestions, and scheduling coordination are seeing the fastest and most comfortable adoption, since a human still reviews the output before it affects a candidate.
More cautious adoption
Automated scoring or ranking of candidates, and anything resembling an automated rejection, is being adopted far more carefully — and in many organisations, deliberately kept as a human decision, with AI used only to prepare the information the decision is based on.
Regulation and human oversight
Regulators are formalising this caution. Under the EU AI Act, AI systems used for recruitment and selection are classed as high-risk, and Article 14 requires that they can be effectively overseen by people. In New York City, Local Law 144 requires an annual bias audit and advance notice to candidates when an automated employment decision tool is used to screen them.
This is general information, not legal advice. Requirements differ by country and change over time, so employers should check the rules that apply to them.